Final Project - GIS 4043

The final presentation for GIS4043 is the final culmination of three-part project: data analysis of the Bobwhite-Manatee 230 kV Transmission Line Right-of-Way.

Step 1 was to review the data, and create a tentative workflow model.
Step 2 was to analyze the data.
Step 3 was to present the data as a powerpoint presentation, and to provide slide notes to go along with it.

These I present here:
http://students.uwf.edu/ear25/Intro2GIS/FINAL_ER.pptx
http://students.uwf.edu/ear25/Intro2GIS/SlideNotes_ER.pdf

This class has been very informative, and I know I learned quite a bit (particularly about those tricky projections!). I enjoyed all the learning challenges, and I know that I will be drawing on lessons presented within this course in the future.

Final Project - GIS 3015

The final task for GIS 3015 was to gather data for 2013 college placement exams (a choice of either ACT or SAT) and present it on a map. The map is intended to be an infographic for a fictional newspaper article in The Washington Post. There were two main objectives with this map. First, the map must be visually appealing, use the best possible thematic data representation, and adhere to basic cartographic design principles. Second, the data presented in the map must be done in the best possible way, with the choice being ours as to what (if any) data classification method to use or which thematic map style(s) to display.

Map infographic depicting 2013 ACT Data.

Technical Details

Graduated picture symbols (representing a mortarboard) were used to depict the composite ACT scores, and a choropleth map was used to represent the percentage of students who took the exam. Since the data focus is mainly on the composite ACT scores, those took first priority (and thus the is the first item visible on the map). The graduated symbols were range graded using manual breaks divided into seven classes – I wanted to preserve the idea of the whole number composite scores, since most people think of the scores in that way. The choropleth data was divided into five separate classes using the quartile method. The percentage dataset was evenly distributed (more or less), with the only ‘outlier’ being the eight states that have mandated 100% participation for high school students. Information on the 2013 ACT was added to an excel table which was later joined to the state shapefile – thus allowing me to use the same shapefile in two different ways on my map.

Since the map is intended to be an infographic, I wanted to catch the reader’s eye and give them something to think about in the few seconds that I have their attention. To do this I steered well clear of any tables on the map – my own eyes glaze over when I see that, and it would require a bit more effort on the reader’s part to make sense of my map (not to impugn the good readers of The Washington Post!). To create visual interest I used a picture symbol for the ACT score data, changed the map neatline to an oval, and arranged my map items accordingly. It is hoped that the deviation from a non-square map might be visually appealing. To further enhance the prominence of the choropleth map (as opposed to the surrounding map text) I used a drop shadow on the contiguous U.S. outline, and also for Alaska and Hawaii.

Final Thoughts on the Class

I found the class to be very beneficial overall. One of the (many) reasons that I had entered into this GIS Certificate program was to learn how to create better maps – what better way to do that than to take on the principals of map design? I found it a relief to learn that I was at least on the right track with my pre-GIS Certificate program maps… and to learn ways to improve on them. It was fun (even in the depths of CorelDraw hell) to try my hand at all those crazy thematic maps – and to pick up some very useful graphic arts knowledge along the way. I know that any maps I will make in the future will be 100% better for having taken this course.

Week 13 - Georeferencing, Editing, & ArcScene

This week marks the last official lab assignment before the final. For this lab we georeferenced aerial photos to buildings and roads associated with the UWF campus, created building polygons and road polylines, and created multiple ring buffers around an eagle's nest. To top it all off, we overlaid data on a DEM in ArcScene - which is a 3D mapping program offered by ESRI.

Georeferenced aerial view of the UWF Campus - and the Eagle Nest location!

Notes on Map 1

The first map shows two separate data frames that contain aerial views of the UWF Campus. The aerial image on the left was georeferenced, and the items in red indicate those features that were digitized during the lab. I've decided to separate out the digitized data from the non-digitized data by color, so that both of my maps sort of flow together.

The data frame on the lower right shows the location of the eagle's nest in relation to the overall boundary of the UWF Campus. The aerial image in the background was provided by ESRI. During the lab I was wondering who in their right mind would ever think building student housing in a wilderness sanctuary would be a great idea... and then I saw that the eagle nest location is right on the university boundary. It sort of makes sense now (but not entirely - I mean, who couldn't see that roadblock coming?).

3D view of a select portion of the UWF Campus.

Notes on Map 2

The second map shows a 3D snapshot view of the UWF Campus, with the digitized buildings 'highlighted' in red. I kept the symbolization and overall map appearance the same as the first map. The buildings were extruded off the surface, and all visible layers are resting on top of a DEM (not immediately visible). Due to the way ArcScene works it was not possible to provide a scale or north arrow... it's almost kind of liberating! Technically the map above is a screenshot of what was visible in ArcScene, with the finishing touches made in ArcMap.

Week 12 - Google Earth

This week in GIS3013 we created maps using Google Earth... or more specifically, created tours using Google Earth.  Data was added from ArcMap by converting our dot map density maps and associated data (from Module 10) to KMZ files.  Afterwards we created a video tour of various urban areas in southern Florida.

A typical street scene of downtown Tampa, as depicted by Google Earth.

Some thoughts...

The main objective with this lab was to familiarize ourselves with Google Earth, and become comfortable adding information generated in ArcMap to the Google Earth program. Converting the maps and individual shapefiles was easy enough, although the difference in output within Google Earth was striking. I found it interesting that there are more options to manipulate the data when it's been converted to KML as an individual layer as opposed to having been imported as group of layers associated with a map.

My biggest struggle with this lab was with the video tour... I mean, that should have been so easy, I'm almost ashamed of how difficult I found it to be! I did my best to keep the view going smoothly - not too fast, not many sudden movements - but that turned out to be for naught when nothing played back the way I thought it would.  During all my video takes I had turned off the southern Florida dot density data (reused from Module 10) during the zoom in to downtown Miami, which was the first stop on our list.  However you'd never know it because that data either wouldn't show up at all during subsequent playbacks, or it would stay turned on the entire time.  What gives?

Another minor snafu was with the importation of the dot density data from Module 10... I had suppressed the mini-dramas involving the dot density mask and so it took a while to figure out why those dots weren't showing up the way I had remembered them.  Plus the legend or the labels wouldn't import either, as they had been converted to graphics/annotation.  Since we had an example dot map included with this lab I used that, but I believe that I could have gotten around my issues by importing the mask as an individual shapefile... and also by completely re-doing the legend and labels on my original map.

Overall I do feel more comfortable adding data to Google Earth and using it generally.  It is an internet needy program though - on my slower connection all the downtown scenes looked like warped views of Gotham city, with small items eventually popping up like bugs once the graphics were loaded... which was kind of neat to me, but probably not in the way the designers intended.  To compensate for this I tried to steer my video tour well clear of too many 'detail' views!

Week 12 - 3D Mapping

This week in GIS3015 we explored the world of 3D mapping with ArcScene and Google Earth. Most of our time was spent with ArcScene within the context of an ESRI online training course. The course provided an overview of how to use ArcScene, how to extrude features, set lighting effects, and much, much more. The course also introduced the concepts of terrain datasets and multipatches - both of which are used extensively within ArcScene to model features and provide a 3D-like experience.

Outside of the ESRI course we created 3D images from 2D features - in our lab example it was the buildings of downtown Boston. The buildings were made using created mass points and building roof elevation values. Our final result was exported as a KML file, for viewing in Google Earth.

Gray 3D buildings add dimension to downtown Boston.

Using 3D maps definitely has a 'wow' factor attached to it, so if attracting an audience is your goal then this is the medium to use. Its analytic capabilities are also pretty cool - some of the examples provided in class included showing varying rates of sun exposure on buildings, or views of available space for rent by floor within individual buildings. One use that really popped out at me is how it can show 'below ground' features and enhance the elevation of a dataset with vertical exaggeration effects. To me this would make a pretty neat way to show archaeological excavations in context. Seriously, people spend hours upon hours analyzing this type of stuff - how awesome would it be to see it fully realized in 3D?

There are some drawbacks to using 3D however - first and foremost being the creation of 3D objects. Essentially, the learning curve can be steep (and costly) when it comes to creating the individual elements that make up a 3D map. For realistic looking data one might need to spend a lot of time and money creating it. And 3D is not always the answer for all situations - its very appearance means that the map can only be viewed on a computer. Sure, there are screenshots that can be taken and printed, but why? If you were going to print out the map anyway then why go through all the trouble of creating one in a specialized 3D format? On the whole though I think that for the average data analysis-minded user the benefits outweigh the costs - especially since for basic analysis one does not need fancy graphics, and learning the basics is no more difficult than learning ArcMap itself.

Week 12 - Geocoding, Network Analysis, and Model Builder

This week in GIS4043 we covered geocoding addresses, used the network analyst extension to create a route, and edited a model in Model Builder. It was quite the whirlwind week!

The map below shows the results of the geocoding and network analyst exercises. The addresses were geocoded from a table of EMS locations within Lake County, Florida. To get the geocoding down we first created an address locator. The points on the map below are the results of the geocoding process (and the subsequent address matching process - which basically involved matching problem addresses using Google Earth and the Lake County EMS provider website).

The route was created from various stop points created at random using the network analyst tool. Basically what is shown below is the optimal route based on pre-assigned conditions (in my case, the best route at 4:30 p.m. on a Monday with U-Turns allowed). The network analyst tool navigated through all of the road data to find the quickest travel times and came up with the route below. If you're interested, the estimated travel time is 11 minutes (stopping at both locations), and the total distance traveled is 8.3 miles.

Geocoded EMS Locations and Optimal Route generated by Network Analyst.

For me the most difficult part of the exercise was the geocoding. It seems so simple - match an address with a location, right? Well, not entirely. The unmatched areas didn't really line up well with reality, and what was even more disturbing is that both Google and Bing maps had serious placement discrepancies. In the end I was left with using Google aerial map view (to look for possible EMS-like buildings) and the Lake County, Florida EMS location webpage.

Both options really messed with my comfort level concerning data accuracy because, well, it's not accurate. However I do realize that if that were my job, I may not EVER be able to go out into the field to ground-truth those address locations. The bigger lesson here was that data entry problems really do create bigger problems down the road...